Research On Transmission Optimization Based On Data Aggregation Problem And Malicious Nodes Detection

In wireless sensor networks, Because in WSNs node has the characteristics of large-scale intensive deployment, similar nodes collect data generally have great similarities, which leads to a lot of information redundancy, so the sink node is not necessary to collect all information. Based on this characteristic, the relevant data collection namely data polymerization process became the main data collection method in the network and the results of the treatment data aggregation fully make up for the shortcomings in WSNs.And in the network transmission process, the network vulnerable to be attacked by attacker, the attacker through deploy the malicious nodes in the network to destroy the network and theft the data of network, caused a lot of damage to the network. So it is necessary to detect the malicious nodes and eliminate the influence of them off.In this paper the above two problems were put forward solution:(1) According to the related data collection, comprehensive consideration of the node hop number, residual energy, polymerization coefficient etc, established a distribution, based on the adaptive adjustment of the dynamic polymerization tree algorithm. According to the average degree of polymerization tree and node energy information dynamic to refresh and maintenance the tree structure, ensure the balance of energy use.(2) Based on the problem of network security we propose a distributed malicious nodes detection algorithm BMND based on Bayesian voting mean. Each node determines its suspected malicious nodes by its request abnormal behavior message. And then determine the malicious nodes by bayesian voting mean. Then the networks can self-heal itself of such malicious nodes influence. The simulation results show the effectiveness and efficiency of our scheme.